JPS5919852A - Method for measuring ultasonic characteristic of bio-tissue - Google Patents

Abstract

PURPOSE:To make regression calculation easy and to determine the characteristic value of bio-tissue from the parameter obtd. by the calculation by normalizing the characteristic function of the bio-tissue that is unaffected by a measurement system and further normalizing the same again by the value of a certain frequency thereby eliminating the influence of the discontinuous transmission at the boundary of the tissue region and taking logarithm by dividing the estimation function by a non-exponential function term. CONSTITUTION:The collected data is fed to a frequency analysis part 7, and the result of the frequency analysis is fed to a data memory 8. The reflection wave spectrum from the bio-tissue as shown by, for example, A and the reflection wave spectrum from a standard reflector as shown by B are obtd. as the result of the frequency analysis. The spectrum A is normalized by the spectrum B in an operation processor 9, by which a spectrum C is obtd. and the frequency corresponding to the coefft. of reflection is determined. The frequency is further regressed by the measurement value obtd. by dividing the same by a non-exponential function term to take logarithm, whereby the parameter associated to the characteristic value of the bio-tissue included in the exponential part of the exponential function is determined.

Description

Detailed Description of the Invention (Technical Field of the Invention) The present invention uses reflected waves of ultrasonic waves, and the attenuation constant is proportional to the frequency (proportionality constant β), such as in living tissue, and the reflection coefficient is different from various living tissue characteristic values. This invention relates to a method for measuring various characteristic values such as β in a medium that is a function of frequency determined by , and particularly relates to a method for determining various characteristic values by regression from power spectrum measurement values to theoretical and experimental formulas.

[Prior art to the invention]

Conventionally, the attenuation constant in the transmission of the fl sound wave is proportional to the frequency f, and the proportionality constant β is a quantity indicating tissue characteristics, and the reflection coefficient is proportional to the n-th power of the frequency f, and this grave index n is It is experimentally known that this is another quantity that indicates tissue properties.

In addition, Ueda et al., the inventor of the present invention, found that the reflection coefficient was

6. Myself. -cd1'no2 It is theoretically shown that the form is Here.

b and d are quantities related to tissue characteristics.

In this case, where the reflection coefficient is a function of f, there is no general method for determining biological tissue characteristic values from the shape of the power spectrum, but the energy ratio method by Miwa et al., the inventor of the present invention, is used.

A patent application has been filed for a method for determining n and β when the functional form is r raised to the nth power.

Although this method is effective, since it focuses on at least three narrow band energy components, it tends to cause errors due to local irregularities in the spectrum, known as so-called scalosoping. Therefore, it is necessary to perform calculations on many sets of three frequencies within the effective band and perform statistical processing on the obtained set of n and β, which has the drawback of requiring a large amount of calculation.

[Purpose of the invention]

An object of the present invention is to obtain a transfer function of a living body from a received signal of an ultrasound reflected wave, and to obtain biological tissue characteristic values.

■A method for determining the shape of a spectrum that avoids discontinuous changes in tissue region edges.

■While the functional form estimated from theory and experiments is expressed as the product of an exponential function term and a non-exponential function term, a method is provided that allows easy regression from actual measured values.

(2) To provide a method for determining the parameters included in the functional form by the regression, and further determining biological tissue characteristic values.

[Structure of the invention]

The present invention normalizes the frequency response spectrum of a biological tissue transfer function, regardless of its absolute value.

In addition to focusing only on its shape, if the functional form estimated from theory or experiment is given by the product of an exponential function term and a non-exponential function term, the measurement obtained by dividing by the non-exponential function term and taking the logarithm. Parameters related to biological tissue characteristic values included in the exponent part of the exponential function term are obtained by regression from the values.

[Embodiments of the invention]

As shown in Figure 1, the biological tissues are distributed from the surface of the body to the deep part of the body (
Ultrasonic pulses (center frequency fo, band 2Ω) are transmitted by the transducer (in direction Z), and the pulse waveforms sequentially reach the internal parts of the body at the speed of sound C, and the reflected waves from each one successively reach the speed of sound C. It propagates in the reverse direction at C and is again received by transducer -2.

Now, suppose that the characteristics of the tissue at depth 2 are to be measured. It is assumed that the living body is composed of i kinds of tissue regions up to a depth of 2. Further, it is assumed that the sound speed C is approximately constant.

The sound pressure pulse transmitted from the body surface (2), ie, 2=0, is attenuated by an attenuation constant αi proportional to the frequency f in each region i. Let this constant of proportionality be ai.

αi=a++βr*f (ai is a constant), and a
i is a parameter indicating characteristics corresponding to tissue i and is abbreviated as attenuation slope.

Furthermore, within each region, a power reflection coefficient r(f) which is a function of frequency f is shown.

In addition, the acoustic impedance changes greatly at the boundaries between each region, and the surface is often approximately smooth, and when passing from this boundary 1, for example, from region i-1 to region i, there is a slight transmission in one step. This results in the loss of one member. Let this transmittance be τ
Let it be j.

The same is true when the reflected wave from depth 2 passes from region i to i-1 in the opposite direction, and this transmittance is defined as τ'i. It may be considered that τi and τ'i have no frequency characteristics.

Now, after transmitting, the transducer ■ successively receives reflected waves from deeper places.

After the reflected wave corresponding to depth 2 is transmitted, t = 2 z /
Since it is received at time C, tissue characteristics at depth 2 can be determined by analyzing the reflected waveform within a certain minute time width τ around that time.

Power spectrum E of the reflected wave received signal from depth 2
r ff) can be easily realized using a well-known frequency analyzer such as FFT.

This E r (f) is the frequency characteristic of the transducer.

It is given by the square of the product of the frequency characteristic of the convergence of the beam at depth 2, a transfer function caused by other measurement systems, and a transfer function of living tissue caused by transmission/reflection in the living body.

Now, if we place a standard reflector underwater at a depth of 2 and find the power spectrum E o [fl of the received wave, Eo (f) is considered to represent the square of the transfer function of the measurement system. . Here, when Er(f) is normalized by dividing it by E o ffl, the square of the transfer function specific to the biological tissue is obtained. The power spectrum of this actually measured biological tissue transfer function is denoted by R (fl).

On the other hand, R(fl should be expressed by a linear equation from the above explanation.

k: a constant that does not depend on f 11: Passage length in region i Below, a method for determining nine tissue characteristics by comparing R(fl with the row of equation (1)) will be described in detail.

One transducer can only calculate R([1) for f within its effective bandwidth, so in order to calculate R(f) over a sufficiently wide range of f, multiple transducers must each It is sufficient to find each R(f) for the bandwidth and connect them.

Furthermore, in actual measurements, the shape of the spectrum exhibits local unevenness due to interference of reflected waves from nearby reflectors and other factors, resulting in the phenomenon of so-called spectral scalosoping, which significantly impairs the measurement of the spectrum shape. lead to errors. To prevent this, measure the front, back, left, right, and upper and lower rings of the measurement point.

Statistical processing such as measuring multiple times and averaging them temporally and spatially is required.

Now, in equation (1), a certain frequency fo (for example, (
1) R(f
When formula (1) is normalized by o).

Then, the factors of τi and τ'i can be removed. Eliminating the unknowns τi and τ′i has a great effect.

Since fo and R(fo) at fo can be found using actually measured values, it is also possible to find the value normalized by R(fo). That is, without being influenced by τi and τ′i, assuming that P(f) and Q If) are equal, Q (f
) can be determined by regression.

Next, r ([There may be various functional forms depending on the degree of approximation of Theoretical formula b Function of f related to transducer dimension 1 curvature (
(known) b = constant σ depending on the average microstructure of the tissue: value of the tissue Spatial autocorrelation distance in the transmission/reception direction C: speed of sound will be explained.

Although equations (4) and (5) have significantly different forms at first glance, they almost match in a certain practical frequency range.

Substituting equation (4) or equation (5) into equation (2), we get
Q4. Q5 is obtained.

Q 4 = A4・fηexp[(-4, gρdJf]
(61Q5-)\,,b'・f':exp[nee4(2gopdt)f-Cshiot2)ze4shi=fr”ノ
(7) As can be seen from equations (6) and (7), Q is characteristically given by the product of an exponential function term and a non-exponential function term. Divide both equations by their respective non-exponential terms and take the logarithm.

A Q4/ A4- '7? −Jnf = −4(Σβ
i Ji) ・f (81J!5QrJ/Ar
,−,lnl;r”=−4(Zβ;lb)・f
-(?) White 9) Here Q4. The parameters n, 'Eo;, ii, ,y
, etc. can be obtained.

Equation (8) has an unknown number n on the left side, but it is sufficient to change n variously to determine n for which the right side is a linear equation.

By determining Σβ77t at each depth Z, βi at each l can be determined by the difference. When β is a continuous function with depth 2, this line integral is Iβd
z, so we can find βfZl by differentiation, or find the X-ray C
It can be calculated from line integrals in each direction using the T algorithm.

In this way, n, β, and l can be determined on one measurement direction scanning line. By scanning this scanning line within a certain plane, it is also possible to obtain a distribution image of n and βt on that plane.

The principle has been described above, but in order to actually implement the device, the technique of extracting parameters from a sample point at a certain depth 2 is widely used in Doppler measurements, etc., and the extracted waveform is subjected to Fourier analysis. As a device for determining the power spectrum, a widely known device such as a DFT (digital Fourier transform device) can be easily used. Its output can be easily input into a calculator and all subsequent calculations can be accomplished by the program. Below is an overview of the system.
This will be explained with reference to FIG.

The timing control unit 1 sends the ultrasound transmission synchronization signal Pd shown in (a) to the transmission circuit 2, and the ultrasound transducer 3 is driven with a pulse having enough power to transmit ultrasound.
Ultrasonic waves are transmitted into living tissue (or into a medium containing a standard reflector).

The reflected wave from the living tissue (or standard reflector) 4 is received by the transducer 3 again, amplified to an appropriate level by the receiving circuit 5, and sent to the data collection unit 6 as the received signal Vr shown in (b). It will be done.

The timing control unit 1 sends the gate pulse pg shown in (c) to the data collection unit 6 at a timing delayed from Pd by a time TI corresponding to the distance from the vibrator surface to the reflection site to be measured.

Capture the reflected signal from the desired area as data.

Note that the width τ of Pg is determined depending on the range to be measured. The collected data corresponds to a waveform as shown in (d), for example.

The collected data is sent to the frequency analysis section 7.

The frequency analysis results are sent to data memory 8.

As a frequency analysis result, for example, a reflected wave spectrum from a living tissue as shown in 5 (A) and a reflected wave spectrum from a standard reflector as shown in 1 (B) are obtained.

The arithmetic processing unit 9 performs various calculations as described above on the frequency analysis results held in the data memory 8.
Get the desired result.

Specifically, the spectrum shown in (A) is normalized by the spectrum shown in (B), the spectrum shown in (C) (corresponding to equation (1)) is obtained, and the rm of equation (3) is , and then calculate the left sides of equations (8) and (9). After that, regression calculation is performed to determine nβ and σ. βi can also be found from β at each depth 2.

The configuration of (2) may be any configuration as long as it can realize the above calculations (2) For example, a microprocessor, RAM, etc.
, ROM, I10 port, etc. may be used.

〔Effect of the invention〕

According to the invention.

■By normalizing, a biological tissue characteristic function that is not affected by the measurement system is extracted, and by further normalizing it with a certain frequency value, the influence of discontinuous transmission at the boundary of one tissue region can be removed.

■By dividing the estimated function by a non-exponential function term and taking the logarithm, one-regression calculation becomes easier.

(2) It has the effect of making it possible to obtain tissue characteristic values from the obtained parameters.

[Brief explanation of the drawing]

Figure 1 is a schematic diagram showing a cross section of a living tissue structure along the Z-axis in the ultrasound scanning direction, where ■ indicates the transducer, ■ indicates the body surface, and i indicates the boundary between area i-1 and area i. , 1 indicates the ultrasound path length in each region. FIG. 2 is a block diagram of a device according to an embodiment of the present invention.
2 is a timing control unit, 2 is a transmitting circuit, 3 is an ultrasonic transducer, 4 is a standard reflector, and 5 is a receiving circuit. The data collection unit 27 is a frequency analysis unit, 8 is a data memory, and 9 is an arithmetic processing unit. FIG. 3 is a diagram showing waveforms at various parts in FIG. 2. FIG. 4 is a diagram showing the relationship of frequency characteristics.

Claims (1)

[Claims] (11) A method for measuring biological tissue characteristics by transmitting ultrasonic pulses into a living body, receiving the reflected waves, and (2) analyzing the received signals. Characteristic ultrasonic biological tissue characteristic measurement method. 1) A step of frequency-analyzing the reflected wave reception signal from the area to be analyzed and obtaining its power spectrum. (1) Normalize the power spectrum obtained in step A above using the power spectrum of the transmitting/receiving measurement system when a standard reflector is provided at the site, and calculate the biological tissue transfer function Rff)
step to find. C) A step of calculating P ffl which is normalized to R([1) using the value R(fo) at a specific frequency fo.2) On the other hand, the biological tissue transfer function predicted from theory and/or experiment is set to the same value (fo The function Q ff) normalized by the value of Q(fl = A
) is divided by fl. P tfl / A (Step of finding fl. E) Step of taking the logarithm of the value obtained in the second step above. F) As a function of the frequency obtained in step E above. Assuming that the value is equal to B ffl in the prediction formula. A (fl, B (regression step to find the optimal value of various parameters included in fl. (2) A (f) in the above second step is the frequency f
B (where fl is expressed as a linear expression of f, and f
2. The biological tissue characteristic measuring method according to claim 1, wherein the first-order coefficient is a quantity related to a line integral of a frequency slope (attenuation slope) of an attenuation constant. (Go to the second step in 311 B (fl is the frequency f
The constant term, the coefficient of the linear term, and the coefficient of the quadratic term are expressed by the average fine structure, attenuation (oblique integral,
The biological tissue characteristic measuring method according to claim 1, wherein the amount is related to an autocorrelation distance in the scanning direction.